Moderate evidence
More
Korean adults distressed without AI felt lonelier over time, a panel study of 1,022 finds
Study title: Generative AI Use and Two-Wave Change in Well-Being and Ill-Being: A Pre- and Post-generative AI Era Panel Study
- Evidence type
- Observational study
- Sample
- n = 1,022
- Sources
- 1 sourcePeer-reviewed
- Published
- 2 min read
- Truvace record
- TRV-2026-3028cite or check it
Study title: Generative AI Use and Two-Wave Change in Well-Being and Ill-Being: A Pre- and Post-generative AI Era Panel Study
A panel survey of 1,022 South Korean adults, surveyed in October 2022 before generative AI spread and again in March 2026, tracked changes in self-esteem, life satisfaction, loneliness, anxiety and depression. Simply adopting generative AI, or the time spent using it each day, was largely unrelated to those changes.
Read the study abstract
As generative artificial intelligence (AI) rapidly spreads, concerns about its impact on users' mental health have surged. However, previous studies have yielded contradicting results owing to theoretical and methodological limitations. To address these gaps, this study investigates how generative AI adoption, usage time, and multidimensional AI dependency are related to two-wave changes in subjective well-being (self-esteem and life satisfaction) and ill-being (loneliness, anxiety, and depression). A two-wave l…
The problem
In a two-wave panel survey of 1,022 South Korean adults (October 2022 and March 2026), withdrawal-type dependency on generative AI, meaning distress or frustration when unable to use it, was associated with less decline in loneliness.
In brief
The exception was withdrawal-type dependency, marked by psychological distress or frustration when people cannot use AI. It was associated with less decline in loneliness over time, and with a suggestive but less robust pattern for depression.
The authors conclude that generative AI use is neither inherently beneficial nor intrinsically harmful. They suggest researchers and practitioners look at maladaptive dependency rather than adoption or usage time alone, treating it as a possible marker of less favorable mental-health change.
Main points
- The study surveyed 1,022 South Korean adults twice: October 2022 (before generative AI spread) and March 2026 (after).
- Outcomes were self-esteem and life satisfaction (well-being) plus loneliness, anxiety and depression (ill-being).
- Multilevel modeling examined within-person and between-person differences at the same time.
- Adopting generative AI and daily usage time were largely unrelated to changes in well-being or ill-being.
- Withdrawal-type dependency was associated with less decline in loneliness.
- The pattern for depression was suggestive but less robust.
- The authors describe dependency as a potential marker of less favorable mental-health change, not a proven cause.
The full story
The researchers say earlier studies of AI and mental health produced contradicting results because of theoretical and methodological limitations. Their design uses a pre-generative AI baseline from October 2022, so change can be measured against a period before the technology became widespread.
They examined three aspects of use: whether people adopted generative AI, how long they used it daily, and multidimensional AI dependency. Only the withdrawal-type dependency, in which people feel distress or frustration when unable to use AI, showed an association with mental-health change.
Because the association was with less decline in loneliness rather than a rise, the finding concerns who improved less over the period. The authors call for future research and practice to examine this maladaptive form of dependency instead of focusing only on adoption or usage time.
Sources
1 on file- Cyberpsychology, Behavior, and Social Networkingdoi:10.1177/21522715261506635
Talk it over
The study found that adopting generative AI and daily time spent using it were largely unrelated to changes in well-being. How should that shape what people watch for in their own AI use?
Korean adults distressed without AI felt lonelier over time, a panel study of 1,022 findshttps://truvace.com/item/generative-ai-use-and-two-wave-change-in-well-being-and-ill-being-a-pre-and-post
The study found that adopting generative AI and daily time spent using it were largely unrelated to changes in well-being. How should that shape what people watch for in their own AI use?
Answer in the threadThe 1,022 South Korean adults reported their own loneliness, anxiety and depression in October 2022 and again in March 2026. How might people's own ratings of distress without AI shape what the survey picked up about dependency?
Answer in the threadThe authors call withdrawal-type dependency a possible marker, not a proven cause. What finding would settle whether distress without AI leads to less decline in loneliness, or the other way round?
Answer in the thread
If you have more time
Which questions would you ask a person to tell withdrawal-type dependency, meaning distress or frustration when unable to use AI, from simply liking a tool?
Answer in the threadThe pattern for depression was suggestive but less firm than the one for loneliness. What kind of follow-up study, and over how many years, would you want to see before anyone acted on it?
Answer in the thread
ace
What readers say